feat: Configure driver to support GPU tests

[orfeas-k]

Implement changes as speced in the spec.
TDLR:

• Split tests to cpu and gpu
• Add --include-gpu-tests flag to include the tests under the gpu directory.
• Pass this flag to the tests job (together with the filter one) to include the appropriate tests.

Closes #132

Testing

In order to test the PR

• Spin up a cluster with a GPU. For a two node microk8s node with one gpu, refer to this comment Exploration task for running UATs using GPU #130 (comment)
• Run with the --include-gpu-tests flag e.g.

  tox -e uats-remote -- --include-gpu-tests --filter "placeholder"
  

  You should see only placeholder in the list (due the flag including it and filter making sure no other tests are run) and the job logs should include:

test_notebooks.py::test_notebook[placeholder] 
-------------------------------- live log call ---------------------------------
INFO     test_notebooks:test_notebooks.py:48 Note that only CPU tests will be run. In order to run tests that use an NVIDIA GPU,use the `--include-gpu-tests` flag e.g. `tox -e kubeflow-local -- --include-gpu-tests`.To learn more, use `--help` or refer to the repository's README file.
INFO     test_notebooks:test_notebooks.py:55 Running placeholder.ipynb...
PASSED                                                                   [100%]

======================= 1 passed, 9 deselected in 7.56s ========================
PASSED

You could also run something without --include-gpu-tests in order to confirm that this PR does not affect running cpu UATs.

[NohaIhab]

thx @orfeas-k , left a couple comments, and:

1. Is the README.md going to be updated in a different PR?
2. Can you provide testing instructions for the PR? Not sure how to tackle this